Mid-cohort check-in

I can structure compound, multi-step prompt chains for real work — not just one-shot prompts.

I deliberately pick the right model + harness (Claude Code / Design / Project, Figma Make / AI) for the task, not by habit.

I can clean a messy backlog using AI — extract themes, dedupe, cluster, map to features — at scale.

I can build a reusable Claude Skill that packages a workflow my team can invoke.

I draft PRDs / specs with AI as a collaborator and arrive at sharper outputs than I would alone.

I can use my Design System as a guardrail when generating with AI — tokens, refusal patterns, the Critic.

I can run an AI prototyping loop that produces DS-clean screens, not chaos I need to clean up.

I know when to reach for Figma Make vs Figma AI vs Claude Design vs Claude Code — and why.

Where are you today on your AI-adoption arc?

Read each stage. Pick the one you'd actually describe yourself as this week — not where you wish you were.

Where has the training helped you most?

Pick the top 3 in order — 1 = biggest help.

One thing you're doing differently this week vs. last.

Anything in the sessions so far you were already doing confidently and could've been skipped or compressed?

The right depth — not too shallow, not too academic.

The right pace.

Enough hands-on time.

The examples and case studies land.

The live demos clarify, not confuse.

PG explains things clearly.

PG responds well to questions and what the room needs.

Energy and engagement in the room are right.

I feel safe to try, ask, and fail in this room.

One real use case from your own work you've started on, even if early.